Bio
I am a Staff Machine Learning & Audio Scientist at Universal Music Group, working in the Music & Advanced Machine Learning Lab (MAML) led by Elio Quinton. My work focuses on machine learning systems that understand and generate music and audio, bridging research in music information retrieval and generative audio modeling.
Current work
At UMG, I develop large-scale audio AI and generative music technologies that move from research into real products and creative tools, including the Sollos algorithms behind Sound Therapy on Apple Music. More broadly, I work on machine learning systems that integrate music information retrieval and generative audio modeling, translating research advances into practical tools for music platforms and creative workflows.
Research background
I completed a PhD in Music Information Retrieval at Utrecht University, supervised by Anja Volk and Remco Veltkamp. My research investigated computational models of variation and subjectivity in musical harmony, including methods for representing multiple chord interpretations and listener-aware harmonic representations.
During my PhD I conducted a research internship at Carnegie Mellon University with Franz Franchetti, and collaborated with Meinhard Müller (International Audio Laboratories Erlangen) on research in music structure analysis and MIR.
Creative and interdisciplinary work
Alongside my research, I have been involved in projects connecting AI research with musical creativity. I served as director of the AI Song Contest, an international initiative exploring AI-assisted music creation, and developed the AI improvisation system for the self-playing piano used in Jameszoo’s Music for 17 Musicians project.
Earlier in my career, I worked on AI projects in computer vision, audio, and music at RTL Netherlands.
Education and music
I hold degrees in Artificial Intelligence from Utrecht University and Sound Design and Music Composition from HKU University of the Arts Utrecht. In addition to my research work, I remain active as a composer and sound artist, with music and audio work broadcast internationally and appearing in selected and nominated works at international film festivals.
Hendrik Vincent Koops
Machine learning for music understanding and generative audio
I co-organised multiple AI Song Contest editions.
Me and others of the AI Song Contest team were featured in this New York Times article
I was a guest editor for the TISMIR special issue on AI and Musical Creativity